iterative optimization
Iterative optimization involves repeatedly adjusting model parameters through cycles of evaluation and improvement, such as with gradient descent or other optimization algorithms, to progressively reduce the loss function.
- Attractive Metadata Attack: Inducing LLM Agents to Invoke Malicious Tools
- Fisher meets Feynman: score-based variational inference with a product of experts
- From Indicators to Insights: Diversity-Optimized for Medical Series-Text Decoding via LLMs
- Image Stitching in Adverse Condition: A Bidirectional-Consistency Learning Framework and Benchmark
- Inference-time Alignment in Continuous Space
- Motion4D: Learning 3D-Consistent Motion and Semantics for 4D Scene Understanding
- Targeted Maximum Likelihood Learning: An Optimization Perspective
- WMCopier: Forging Invisible Watermarks on Arbitrary Images